Electronic Theses and Dissertations

Date of Award

1-1-2024

Document Type

Thesis

Degree Name

M.S. in Engineering Science

First Advisor

John Daigle

Second Advisor

Lei Cao

Third Advisor

David Harrison

School

University of Mississippi

Relational Format

dissertation/thesis

Abstract

The objective of this thesis is to assess the performance of a data-drivennetwork scheduler in optimizing network performance in a three-slice 5G wireless cellular network. Each slice represents a distinct 5G traffic category or use case. These slices operate on the same underlying physical core and radio network; how- ever, they appear as independent networks from the end-user perspective. This work models the resource scheduling problem as a Markov decision process. A dataset of key performance measurements of a 5G network is constructed using MATLAB to train a deep Q-network reinforcement learning agent. Simulation results show that the agent can make scheduling decisions based on the performance of user endpoints (UEs). However, the reinforcement learning agent did not outperform the traditional round-robin scheduler in ensuring that all UEs achieve their performance objectives.

Share

COinS
 
 

To view the content in your browser, please download Adobe Reader or, alternately,
you may Download the file to your hard drive.

NOTE: The latest versions of Adobe Reader do not support viewing PDF files within Firefox on Mac OS and if you are using a modern (Intel) Mac, there is no official plugin for viewing PDF files within the browser window.